Cargando…
Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat?
The interpretability of deep neural networks has attracted increasing attention in recent years, and several methods have been created to interpret the “black box” model. Fundamental limitations remain, however, that impede the pace of understanding the networks, especially the extraction of underst...
Autores principales: | , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
John Wiley and Sons Inc.
2022
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9762288/ https://www.ncbi.nlm.nih.gov/pubmed/36216585 http://dx.doi.org/10.1002/advs.202204723 |
_version_ | 1784852832818561024 |
---|---|
author | Xu, Hao Chen, Yuntian Zhang, Dongxiao |
author_facet | Xu, Hao Chen, Yuntian Zhang, Dongxiao |
author_sort | Xu, Hao |
collection | PubMed |
description | The interpretability of deep neural networks has attracted increasing attention in recent years, and several methods have been created to interpret the “black box” model. Fundamental limitations remain, however, that impede the pace of understanding the networks, especially the extraction of understandable semantic space. In this work, the framework of semantic explainable artificial intelligence (S‐XAI) is introduced, which utilizes a sample compression method based on the distinctive row‐centered principal component analysis (PCA) that is different from the conventional column‐centered PCA to obtain common traits of samples from the convolutional neural network (CNN), and extracts understandable semantic spaces on the basis of discovered semantically sensitive neurons and visualization techniques. Statistical interpretation of the semantic space is also provided, and the concept of semantic probability is proposed. The experimental results demonstrate that S‐XAI is effective in providing a semantic interpretation for the CNN, and offers broad usage, including trustworthiness assessment and semantic sample searching. |
format | Online Article Text |
id | pubmed-9762288 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2022 |
publisher | John Wiley and Sons Inc. |
record_format | MEDLINE/PubMed |
spelling | pubmed-97622882022-12-20 Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? Xu, Hao Chen, Yuntian Zhang, Dongxiao Adv Sci (Weinh) Research Articles The interpretability of deep neural networks has attracted increasing attention in recent years, and several methods have been created to interpret the “black box” model. Fundamental limitations remain, however, that impede the pace of understanding the networks, especially the extraction of understandable semantic space. In this work, the framework of semantic explainable artificial intelligence (S‐XAI) is introduced, which utilizes a sample compression method based on the distinctive row‐centered principal component analysis (PCA) that is different from the conventional column‐centered PCA to obtain common traits of samples from the convolutional neural network (CNN), and extracts understandable semantic spaces on the basis of discovered semantically sensitive neurons and visualization techniques. Statistical interpretation of the semantic space is also provided, and the concept of semantic probability is proposed. The experimental results demonstrate that S‐XAI is effective in providing a semantic interpretation for the CNN, and offers broad usage, including trustworthiness assessment and semantic sample searching. John Wiley and Sons Inc. 2022-10-10 /pmc/articles/PMC9762288/ /pubmed/36216585 http://dx.doi.org/10.1002/advs.202204723 Text en © 2022 The Authors. Advanced Science published by Wiley‐VCH GmbH https://creativecommons.org/licenses/by/4.0/This is an open access article under the terms of the http://creativecommons.org/licenses/by/4.0/ (https://creativecommons.org/licenses/by/4.0/) License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. |
spellingShingle | Research Articles Xu, Hao Chen, Yuntian Zhang, Dongxiao Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? |
title | Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? |
title_full | Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? |
title_fullStr | Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? |
title_full_unstemmed | Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? |
title_short | Semantic Interpretation for Convolutional Neural Networks: What Makes a Cat a Cat? |
title_sort | semantic interpretation for convolutional neural networks: what makes a cat a cat? |
topic | Research Articles |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9762288/ https://www.ncbi.nlm.nih.gov/pubmed/36216585 http://dx.doi.org/10.1002/advs.202204723 |
work_keys_str_mv | AT xuhao semanticinterpretationforconvolutionalneuralnetworkswhatmakesacatacat AT chenyuntian semanticinterpretationforconvolutionalneuralnetworkswhatmakesacatacat AT zhangdongxiao semanticinterpretationforconvolutionalneuralnetworkswhatmakesacatacat |